Cryopreserved Umbilical Cord Mesenchymal Stem Cells Show Comparable Effects to Un-Cryopreserved Cells in Treating Osteoarthritis

Abstract

Non-cryo and hypothermic preservations are two available options for short-term storage of living cells. For long-term cell storage, cryopreservation is an essential procedure as it prolongs the storage time, allowing for the transport and testing of cells, as well as the establishment of cell banks. But it is unclear whether cryopreservation reduces the therapeutic effects of human umbilical cord mesenchymal stem cells (hucMSCs) on osteoarthritis (OA). To investigate this, we compared the basic biological characteristics and the anti-OA efficacy of un-cryopreserved hucMSCs (UC-MSCs) and cryopreserved hucMSCs (C-MSCs). A mono-iodoacetate-induced rat OA model was established to evaluate the anti-OA properties of UC-MSCs and C-MSCs. And the conditioned medium of UC-MSCs (UC-CM) and cell freezing medium of C-MSCs (C-CFM) were collected for the mechanism study. No significant differences were found between UC-MSCs and C-MSCs in cell viability, immunophenotype, and trilineage differentiation capacity. In vivo, UC-MSCs and C-MSCs exhibited similar cartilage-repairing effects by attenuating pain and alleviating pathological changes in OA rat joints. In vitro, C-CFM and UC-CM promoted the proliferation of chondrocytes, improved the expression of anabolism-related molecules (Col2, COL2, and SOX9), and decreased the expression of catabolism-related molecules (Adamts5, Mmp13, Il6, COL10, and MMP13). These results indicated that UC-MSCs and C-MSCs had comparable anti-OA effects, and cryopreservation did not alter the anti-OA capability of hucMSCs, which provides further support for clinical use of C-MSCs in treating OA.

Graphical Abstract

Human umbilical cord-derived mesenchymal stem cells (hucMSCs) have offered a new option for the treatment of osteoarthritis (OA) ¹ . Although numerous studies have shown the anti-OA efficacy of hucMSCs, most of these studies have used freshly cultured cells^2,3. The application of freshly cultured cells often lacks industrial production flexibility and long-distance transportation practicability, which hardly meet the needs of clinical and translational research. To address these challenges, non-cryo and hypothermic preservations have emerged as feasible solutions, typically for the short-term storage of living cells^4,5. For long-term storage of cells, cryopreservation is considered to be an essential procedure for logistical delivery and cell banking of mesenchymal stem cell (MSCs) ⁶ . Cryopreservation is currently an effective method to store cells for extended periods of time, and allogeneic cryopreserved MSCs have been used in the majority of human clinical trials^{7 –9}. However, many clinical trials using cryopreserved MSCs failed to meet the primary clinical endpoint and their data did not support the clinical use of cryopreserved MSCs, raising doubts upon the efficacy of cryopreserved MSCs products^{10 –12}. A recent study demonstrated that cooling at 4℃ alone could produce extensive DNA damages in cultured cells through oxidative stress and adenosine triphosphate (ATP) depletion ¹³ . Another study has reported accumulated DNA damages in cryostored stem cells in liquid nitrogen resulted from natural radiations ¹⁴ . Therefore, it is necessary to carry out preclinical studies to confirm the therapeutic efficacy of cryopreserved MSCs toward OA.

Here we compared the manifestation of cell viability, growth status, immunophenotype, and trilineage differentiation capability between un-cryopreserved hucMSCs (UC-MSCs) and cryopreserved hucMSCs (C-MSCs) for 6 months. Both UC-MSCs and C-MSCs showed spindle-shaped fibroblast-like morphology and adherent growth state (Supplementary Fig. S1A). The cell viability of C-MSCs was slightly lower than UC-MSCs after culturing for 24 h (Supplementary Fig. S1B). Immunophenotyping histogram showed that both UC-MSCs and C-MSCs expressed surface markers CD73 (>95%), CD90 (>98%), CD105 (>95%), while not HLA-DR (<2%), CD11b (<2%), CD19 (<2%), CD34 (<2%), nor CD45 (<2%) (Supplementary Fig. S1C). According to the specific staining results, both C-MSCs and UC-MSCs had the ability of osteogenesis, chondrogenesis, and adipogenesis differentiation (Supplementary Fig. S1D). The above results indicated that C-MSCs and UC-MSCs had similar expression pattern of surface markers and trilineage differentiation ability.

Next, a mono-iodoacetate-induced rat OA model was established, and the anti-OA properties of UC-MSCs and C-MSCs were evaluated by pain behavioral tests, histopathological and immunohistochemical analysis. A schematic diagram for the entire animal study was shown in Supplementary Fig. S2. The mechanical withdrawal threshold (MWT) and the thermal withdrawal latency (TWL) of OA rats were measured at weeks 2 (0 day after injection) and 6 (28 day after injection) of the whole experimental periods. After OA modeling, both MWT and TWL values of rats decreased significantly (all P < 0.05 vs. Sham), and the status had no obvious change in Model-UC group and Model-C group after 28 days (all P > 0.05 vs. 0 d) (Fig. 1A). However, both mechanical and thermal pain threshold parameters of rats in UC group, C-M group, and C-H group were significantly improved (P < 0.05 vs. 0 d) (Fig. 1A). Thus, C-MSCs and UC-MSCs could comparably relieve joint pain of OA rats.

Figure 1.

Comparable anti-osteoarthritic (OA) effects between un-cryopreserved and cryopreserved hucMSCs. (A) Mechanical withdrawal threshold (MWT) and thermal withdrawal latency (TWL) of rats from different groups measured before starting treatment (0 day) and after the final injection of hucMSCs (28 day). Both un-cryopreserved and cryopreserved hucMSCs significantly increased MWT and TWL values, indicating attenuation of joint pain in OA rats. MWT and TWL testing results were calculated as the average of both paws. Values are presented as mean±standard deviation (n=10/group). ^##P < 0.01 vs. Sham on day 28; *P < 0.05 or **P < 0.01 vs. Model-UC or Model-C on day 28. (B) Statistical analysis of COL2 and MMP13 expressions, and Mankin’s scores for histopathological observation. Both un-cryopreserved and cryopreserved hucMSCs significantly improved histopathological abnormalities and cartilage matrix turnover of knee joints in OA rats. Values are presented as mean±standard deviation (n=3/group). ^##P < 0.01 vs. Sham; *P < 0.05 or **P < 0.01 vs. Model-UC or Model-C. Model-UC: injection of saline after OA modeling. Model-C: injection of cell freezing medium after OA modeling. C-L: low-dose cryopreserved hucMSCs reconstituted in cell freezing medium. C-M: medium-dose cryopreserved hucMSCs reconstituted in cell freezing medium. C-H: high-dose cryopreserved hucMSCs reconstituted in cell freezing medium. UC: un-cryopreserved hucMSCs reconstituted in saline. (C) Cell viability of chondrocytes after treatment of UC-CM and C-CFM for 24 h, with Cho-CM and Cho-CFM as basal culture condition for comparison. Cho-CM, Cho-CFM, UC-CM, and C-CFM groups were divided into two subgroups to investigate the influence of TNF-α on the cell viability of chondrocytes. Both C-CFM and UC-CM increased the viability of chondrocytes, regardless of the presence of TNF-α. (D) The relative mRNA expressions of OA pathological genes in chondrocytes with treatment of Cho-CM, Cho-CFM, UC-CM, and C-CFM. Both C-CFM and UC-CM restored the mRNA expressions of Adamts5, Il6, Col10, Mmp13, and Col2. (E) Protein expressions of OA pathological molecules in chondrocytes with treatment of Cho-CM, Cho-CFM, UC-CM, and C-CFM. Both C-CFM and UC-CM restored the protein expressions of COL2, COL10, SOX9, and MMP13. Values are presented as mean±standard deviation (n=3/group). ^#P < 0.05 or ^##P < 0.01 vs. NC; *P < 0.05 or **P < 0.01 vs. model (TNF-α or Cho-CFM or Cho-CM). C-CFM and Cho-CFM denote the cell freezing medium of cryopreserved hucMSCs and chondrocytes, respectively. UC-CM and Cho-CM denote conditioned medium of un-cryopreserved hucMSCs and chondrocytes, respectively. NC: normal control. TNF-α: inflammation modeling using 10 ng/mL TNF-α.

The histopathological staining results showed irregular cartilage surface, disorganized structure, decreased chondrocytes, and loss of collagen in joints from the model group (Supplementary Fig. S3). While the articular cartilage degeneration was alleviated by UC-MSCs and C-MSCs, and the number of chondrocytes and mass of collagen were obviously increased, especially in the C-H group and UC group (Supplementary Fig. S3). The Mankin’s scoring demonstrated that compared with the corresponding model groups, the histopathological scores of the C-M group (P < 0.05 vs. Model-C), C-H group (P < 0.01 vs. Model-C), and UC group (P < 0.01 vs. Model-UC) significantly decreased in a descending order, indicating their anti-OA effects in the following order: UC > C-H > C-M > C-L (Fig. 1B).

COL2 and MMP13 are key markers of cartilaginous matrix anabolism and catabolism, and their expressions were observed by immunohistochemical staining. As shown in Supplementary Fig. S4 and Fig. 1B, the expression of COL2 was significantly decreased in the two model groups (P < 0.01 vs. Sham). Nevertheless, UC-MSCs and C-MSCs administrations obviously improved the COL2 expression, especially in the C-M group (P < 0.01 vs. Model-C), C-H (P < 0.01 vs. Model-C), and UC group (P < 0.01 vs. Model-UC). As shown in Supplementary Fig. S4 and Fig. 1B, MMP13 was highly expressed in two model groups (P < 0.01 vs. Sham). The administrations of UC-MSCs and C-MSCs decreased the number of MMP13-positive cells, especially in the C-H group (P < 0.05 vs. Model-C) and UC group (P < 0.05 vs. Model-UC). In sum, the histopathological abnormalities as well as anabolic and catabolic balance of cartilage matrix were restored by C-MSCs in a dose-dependent manner.

Considering the paracrine effect is an important mechanism of MSC therapy for OA ¹⁵ , we evaluated the influences of the conditioned medium (CM) from UC-MSCs and cell freezing medium (CFM) from C-MSCs on chondrocytes. The proliferative effects of UC-CM and C-CFM on primary chondrocytes were evaluated by CCK8 assay. As shown in Fig. 1C, neither UC-CM nor C-CFM had cytotoxicity on chondrocytes. Both UC-CM and C-CFM at 1/10 dilution rate significantly increased the viability of chondrocytes after 24 h treatment (P < 0.05 or P < 0.01 vs. NC), with the proliferation rate reaching to 7.62% and 30.44%, respectively. Under the same condition, Cho-CFM and Cho-CM did not promote the proliferation of chondrocytes (P < 0.01 vs. NC). Compared with NC group, the chondrocytes treated with tumor necrosis factor-alpha (TNF-α) had no significant change in cell viability (P > 0.05 vs. NC). The proliferation rate of TNF-α-stimulated chondrocytes was improved by C-CFM and UC-CM treatment (P < 0.05 or P < 0.01 vs. TNF-α), while it was decreased in Cho-CFM and Cho-CM groups (P < 0.01 vs. TNF-α). Therefore, both C-CFM and UC-CM demonstrated beneficial effects on the viability of chondrocytes, regardless of the presence of TNF-α.

In addition, qPCR and Western blot (WB) analyses were conducted to investigate the molecular actions of UC-CM and C-CFM on chondrocytes. As shown in Fig. 1D, the relative mRNA expressions of Adamts5, Il6, Col10 and Mmp13 were significantly up-regulated and that of Col2 was significantly down-regulated in the Cho-CFM and Cho-CM groups (P < 0.05 or 0.01 vs. NC). The abnormal expressions of these genes were significantly restored after a 24 h treatment of UC-CM (P < 0.05 or 0.01 vs. Cho-CM) or C-CFM (P < 0.05 or 0.01 vs. Cho-CFM). As shown in Fig. 1E, the protein levels of MMP13 (up-regulated), COL10 (up-regulated), COL2 (down-regulated), and SOX9 (down-regulated) were significantly altered in the Cho-CFM and Cho-CM groups (P < 0.05 or 0.01 vs. NC). In contrast, UC-CM (P < 0.05 or 0.01 vs. Cho-CM) and C-CFM (P < 0.05 or 0.01 vs. Cho-CFM) significantly restored the abnormal expressions of these proteins toward normal levels. Together, the above results indicated that CM from C-MSCs and UC-MSCs could comparably regulate OA pathological mRNA and protein expression in chondrocytes.

This study has some limitations to consider. One is that we only observed the effects of short-term cryopreservation on the functional properties of hucMSCs, and the effects of long-term cryopreservation warrant further evaluation. Second, although we confirmed that hucMSCs may play an anti-OA role through a paracrine mechanism, the components of CFM and CM were not analyzed, and it is unclear which bioactive component plays the dominant role and whether there is synergy within the secretome. Moreover, we did not investigate the effects of different proportions and types of cryoprotectants on hucMSCs, and it is unknown whether different CFM have an effect on the efficacy of hucMSCs for OA treatment.

To conclude, cryopreservation exhibited minimal impact on the biological properties (cell morphology, immunophenotype, and differentiation capacity) and the anti-OA effects of hucMSCs. Similar to cultured hucMSCs, C-MSCs could protect cartilage by regulating the inflammatory microenvironment and restoring the balance of synthesis and catabolism of chondrocytes. Thus, C-MSCs could be considered as a viable alternative to cultured hucMSCs for OA treatment.

Supplemental Material

sj-docx-1-cll-10.1177_09636897241297631 – Supplemental material for Cryopreserved Umbilical Cord Mesenchymal Stem Cells Show Comparable Effects to Un-Cryopreserved Cells in Treating Osteoarthritis

Supplemental material, sj-docx-1-cll-10.1177_09636897241297631 for Cryopreserved Umbilical Cord Mesenchymal Stem Cells Show Comparable Effects to Un-Cryopreserved Cells in Treating Osteoarthritis by Bo Yan, Huixin Chen, Li Yan, Qiang Yuan and Le Guo in Cell Transplantation